Re: 4.01 panic/lock issues?

On Sat, Jan 10, 2009 at 10:25:01AM +0900, Izumi Tsutsui wrote:
> According to the Sun3 Archive page:
> http://www.sun3arc.org/hardpatches/3.60-tuning/fpu.phtml
> 68882 is software compatible with 68881 (except CPI),
> so I wonder if it could cause FPU related errors.
> Actually I used 68882 on 3/60 about 13 years ago (with 1.1 or 1.2?)
> and it worked fine with X11 etc.

Yeah - and this is why I probably never caught this before. This
specific box has been running 2.0+ for years without issue... I only
started noticing oddities once I got 4.0.1 on it. Everything I'd read,
also, had said that it's full compatible and also faster... the last
part I've wondered about, however, since compared to other sun3s I've
used, this one doesn't seem to really react any faster on fpu-intensive
tasks.

> Hmm, does only the mutt binary cause the problem?

So far, yes - the first time I'd loaded this box, I was having problems
with most packages off of the ftp site. Now that the box has been
reloaded twice since, it appears to be just mutt.  I'm compiling mutt out 
of pkgsrc right now to compare, and also trying to see if I can find any
other apps that will cause the problem.

> How did you install it? From packages in ftp.NetBSD.org?
> If so, in which path?

Correct.

# env |grep PKG_PATH
PKG_PATH=ftp://ftp.netbsd.org/pub/NetBSD/packages/current-packages/NetBSD/m68k/4.0_2008Q2/All

> What command (including option etc.) did you try when the panic happens?
> I.e. how can we reproduce it with a certain procedure?

It happens consistently when I login to a IMAP server:

% mutt -f imap://mail.test.com

After login, it gathers message headers; when it begins to sort the
mailbox (user sees "Sorting mailbox..."), it immediately panics. For
reasons I don't fully understand, re-sorting local mailboxes don't cause
the panic.

> I'm not sure unimlemented FPU trap could happen even with 68882.
> One possibility is binaries compiled with -m68040 or 060 option,
> but kernel shouldn't panic even in that case.
> If you build a kernel without options FPU_EMULATE, I guess
> it won't panic but shows "no floating point support",
> as per src/sys/arch/sun3/sun3/trap.c.

Would it make sense, for troubleshooting-purposes, to recompile without
FPU_EMULATE?